Gemma 4 26B A4B vs MiniCPM5-2B
Wins 4 of 6 areas
Coding · Reasoning · Long documents · Following instructions
Wins 1 of 6 areas
Agents
Gemma 4 26B A4B is the stronger all-rounder.MiniCPM5-2B is better at agents.
Scores updated · 21 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software30Gemma 4 26B A4B3 of 3 tests
- ReasoningHard problems that need careful thinking20Gemma 4 26B A4B2 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts10Gemma 4 26B A4B1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemma 4 26B A4B1 of 1 test
- AgentsCarrying out multi-step tasks on its own12MiniCPM5-2B2 of 3 tests
- FactsGetting facts right instead of making them up11Even1 each
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemma 4 26B A4B pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+30.4points ahead
- Multi-step tasks using many real software toolsMCP Atlas+23.6points ahead
- Code for real scientific research problemsSciCode+13.7points ahead
Where MiniCPM5-2B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+64.5points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+8.8points ahead
- Real work tasks from 44 professionsGDPVal+7.1points ahead
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemma 4 26B A4B
- Terminal-Bench 2.1Gemma 4 26B A4B by 30.4398.6+30.4
- SciCodeGemma 4 26B A4B by 13.74026.3+13.7
- LiveCodeBench v6Gemma 4 26B A4B by 6.677.170.5+6.6
AgentsMiniCPM5-2B
- MCP AtlasGemma 4 26B A4B by 23.65026.4+23.6
- τ-Bench V3 · BankingMiniCPM5-2B by 8.81220.8+8.8
- GDPValMiniCPM5-2B by 7.13.310.4+7.1
ReasoningGemma 4 26B A4B
- Humanity's Last ExamGemma 4 26B A4B by 10.419.38.9+10.4
- GPQA DiamondGemma 4 26B A4B by 979.270.2+9
- CritPttie00.3tie
FactsEven
- AA-Omniscience · Non-hallucinationMiniCPM5-2B by 64.513.678.1+64.5
- AA-Omniscience · AccuracyGemma 4 26B A4B by 10.719.18.4+10.7
Following instructionsGemma 4 26B A4B
- IFBenchGemma 4 26B A4B by 10.372.462.1+10.3
Other results8 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceMiniCPM5-2B by 39.2-50.8-11.7+39.2
- Artificial Analysis Coding IndexGemma 4 26B A4B by 24.839.314.5+24.8
- WebArena-VerifiedGemma 4 26B A4B by 16.525.69.1+16.5
- HLE (with tools)Gemma 4 26B A4B by 5.717.211.5+5.7
- BFCLv4MiniCPM5-2B by 5.657.162.6+5.6
- AA IntelligenceGemma 4 26B A4B by 4.216.712.5+4.2
- IFEvalGemma 4 26B A4B by 4.29287.8+4.2
- AA Agentic IndexGemma 4 26B A4B by 2119+2
Questions people ask
Which is better, Gemma 4 26B A4B or MiniCPM5-2B?
Gemma 4 26B A4B wins four of the six areas where both have results: coding, reasoning, long documents and following instructions. MiniCPM5-2B wins agents. They are level on facts.
Which is better for coding?
Gemma 4 26B A4B. It wins 3 of the 3 coding tests both models report; MiniCPM5-2B wins none.
How do you compare the two?
We use the 21 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 8 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.